Any business owner or marketing specialist wants higher conversion rates with minimal expenses. They try using different techniques, like creating consumer profiles and customer segmentation, and all with one goal in mind: for a customer to visit their salon or website and think, “Wow! That’s exactly what I need!” – and purchase at once.
Many world-famous corporations wrestle with this task, but none of them do it better than Uncle Joe.
Getting 100% Conversion Rate: Uncle Joe’s Case
Every small city has an Uncle Joe. He’s the owner of a small shop where he sells food, newspapers, and some anything-you-might-need stuff. If you come into his shop and watch, you’ll see that almost nobody leaves without a making a purchase.
Uncle Joe knows every single one of his customers. He knows that Henry’s wife caught a light flu two days ago. If Henry comes in, Joe will ask him about his wife’s health and offer him some fresh honey to buy. He knows that Mary has two children, and he has nice T-shirts with their favorite cartoon heroes at a nice discount.
And the conversion rates of Uncle Joe’s small shop can be as high as a crazy 100%. Any corporation would gladly pay all the money in the world for a way to hit the same numbers.
And yes, the bigger the company is, the harder this task becomes. It’s almost impossible for the biggest ones. At least, that’s how it was some time ago, before adaptive marketing came on the scene.
What is adaptive marketing for branding?
At TemplateMonster, a provider of website templates for content management systems (CMS), we had a lot of traffic from India. Many visits, a high number of pageviews, but almost zero purchases. We tried giving them a 50% discount, but that didn’t increase conversion rates at all. Then we gave them a 70% discount, and it brought us 2 purchases. Then we made an India Day on Template Monster and gave them a 90% discount… This got us a whole 4 purchases.
Then I traveled to India. I had to understand what exactly was going on. Do you know what was the first main problem? They just didn’t understand the prices in dollars. All the discounts meant nothing to them, because they knew only rupees. The second problem I discovered was the language.English in India is different from what we use, so the website was difficult for them to read and understand.
Adaptive marketing means making personalized offers to each of your market segments in order to maximize conversion rates. Just like Uncle Joe, who makes different offers to his customers.
Adaptive marketing examples? Here is the one. If you sell computers, you will have a different approach to selling to a parent than to a system administrator. A parent looking to buy a computer for their child would be interested in one that’s not too expensive and suitable for educational purposes. Telling them about RAM or any other specs would be pointless; they just won’t understand your words. On the other hand, a system administrator would find you unprofessional if you didn’t tell them about product specifications.
If you have at least 10 different segments in your target audience, this may seem difficult. If you have more, it may seem impossible to find out how to make different offers to so many different people — unless you use data mining.
How data mining works
First of all, you need a big data set to analyze. You need thousands of people belonging to your potential target audience and tens of parameters like age, sex, buying habits, IP address, browser language, and more. All this data should be gathered and analyzed using a specialized program.
The result of the analysis must be a number of predictive models which are assumptions – for example, which people are more likely to buy the product, or which conditions may raise the probability of conversion.
Then, all you need to do is to test the most promising predictive models and implement them into your marketing.
How the big guys do it
IBM had a similar case. Their website had several million users, but their call center could support only a couple of hundred calls at a time. They faced a typical business dilemma – the one where you have a huge data set, but limited resources. What they wanted was to improve sales without increasing support expenses.
What was the solution? They categorized their website, and used client segmentation to mark different sections of the website based on the visitor’s purpose. For example, there were students among their website visitors who could become their customers years later; engineers who needed technical documentation; top employees of IT companies who might bring the IBM company profitable purchases, etc.
Also, there were certain parameters like IP addresses, page visit history and others that helped them identify which visitors belonged to which category. What IBM did next was to make different offers to different types of customers.
So, if it was an engineer visiting their website, they were directed to the specs and other technical information. If it was a student, probably looking for educational opportunities, they gave him or her other relevant content. If it was their ICP (ideal customer profile), they invited that person to contact their consultants right away.
What was the result? A whopping 350% sales growth, with no additional resources needed. Fantastic, right?
Going back to Indian case
So what did we do in our case? Of course, we changed the text and the currency on the website version for Indian visitors. But that was only the tip of the iceberg.
We followed the same principles described above and found out that Indian designers got inspiration from our templates and went on to make their own. They visited our website without intending to buy anything.
So we adapted. Instead of trying to sell templates, we started offering them opportunities. Education is very important in India, so we offered them courses and certification programs. Another opportunity was to create partnerships.
Ideally, we could take thousands of Indian eCommerce websites as a dataset and learn more patterns of behavior. We had room to grow.
Many companies have already started using adaptive marketing strategies in their own way, and it’s not only in their online efforts.
Take Coca-Cola, for instance. They’ve tried to create different flavors of Coke many times, but none of them stick. People prefer the classics — Coke, Diet Coke, and Zero. But the company still finds ways to engage customers and uphold its brand identity. For example, in 2009, it launched the Freestyle vending machine, where you can mix different Coca-Cola brands of beverages. Now, you can even share your recipe with friends on Facebook, so they can try the same mix. It might not increase Coke’s revenues by much, but it’s certainly fun.
Nike has an exceptional case. They launched NikeID in 2005 before adaptive marketing was even a thing. Today, they ship about 2 million pairs of shoes a day. The idea is that you can customize your running shoes down to the smallest detail, like a colored stripe on the side, and save the ID to re-order the same pair. Imagine that! Their technology is so advanced that they can ship many pairs of personalized shoes every day.
Image source: nike.com
Adaptive content online. How to personalize the data for every user?
Right now, the concept of smart websites is starting to take over the Internet. These websites can predict what the customer is going do on the website.
For example, what can you say about a user who is looking at a cinema website from a smartphone? Most likely he or she wants to do one of three things: check when a movie is showing, check the location, or quickly buy a ticket. So the mobile version of the website should first show the screening schedule, an option to book online, and an address that directs the user to Google Maps. On the other hand, you can predict that a desktop user may want to see the trailer.
So, with smart websites, you’re trying to predict different user cases. And you can do even better. You can show users different content based on their predicted occupation.
How can you do that? For example, business desktops have a typical resolution, so you can predict that this customer works at an office and communicate with them accordingly. Computers for entertainment and design have higher resolutions, so you can also adapt the content you show to this segment.
This way, you can offer personalized content and positively influence your conversion rate. Sounds awesome, right?
Data personalization in practice
Sure, it all sounds great, but where do you start if all you have is just another simple WordPress website? How do you use adaptive marketing in this case?
I suggest you create an experience map. It’s basically your vision of the route, or flow, that each segment of customers should take through your website. For example, B2B and B2C customers would have different flows, since they visit your website for different purposes.
As soon as you draw this flow, you will see how you can adapt your website. A good example is the Rail Europe Experience Map.
It’s comes down to details: how do customers research and plan where they are going to buy their tickets, how do they look for availability, how do they pay for the tickets and all the other steps in the process. Lastly, the map shows how the customer can get any kind of assistance.
Let’s agree that it is almost impossible to keep all this information in mind, so it’s better to keep it on paper or on your computer. Even a simple, single-page website can have an extensive experience map when you draw it. You don’t even need any special skills, just common sense, and a basic understanding of who your customers are and what you want from them.
Now, the only thing other you need to personalize your website is your customers’ personal data.
Where do you get the data?
We can say that adaptive marketing is the active use of personal data with a focus on consumer interests and demands. You can get the right data from:
- Web analytics reports;
- Client surveys;
- Quality one-on-one client interviews;
- Feedback on your company’s blog;
- Social media analytics;
- Big data.
The simplest thing you can do is talk to the customers yourself. If you have a call center or a chat, take about 3 days and have 10 to 20 conversations to conduct deep interviews. Just don’t talk about the product, talk about the client. Get to know your client, find out where they are from, who they are, how they search for information, what their previous experience with similar products was, etc.
Goals and objectives: set up the process for your product
There’s nothing complex about using adaptive marketing when you have a dataset and big-data tools. If anything, most of the work will be done by the computer. It doesn’t even matter what tools you use as long as you understand, conceptually, what your final results should be.
Start with data mining to determine how many ICPs you have. Data science gives you a huge advantage over any knowledge you or the best marketing specialists could ever offer. Let’s say a combination of these conditions gives you a high conversion rate: a Beverly Hills IP address, a certain page depth, and a certain search query. You would see it as a logical chain, while data mining can show you every potential segment.
The task of machine learning is to process all of your potential segments and study whether they are valid or not. Maybe only 50 out of 2000 are valid, but the program will give you concrete segments that can work in real life.
The next step is to form predictive models. They are formed and tested automatically by the program. Here’s an example of such a model: a client visits your website, searches for something, and goes four pages deep; if you show this client that you have an offer today, there is a 70% likelihood of a conversion.
Another automated process is the clustering of customers. The program puts clients into groups based on previously collected information about them.
Up to this point, everything is done by a computer, but ad units require human intervention. You need to form a database of messages for communication with the customers. Basically, it’s your marketing mix.
Now we have only two stages of this process left: behavioral units and communication units.
Each behavioral unit is what we predict a customer will do after seeing the offer. For example, they will type in their email address to receive the discount, or contact the call center. When you have a big database of ad units, at this stage the program can cross-reference them with the clusters of customers and find the optimal combination. It will show you which messages should be shown to which customers. After that, you may even learn how to treat your clients to make them lifelong advocates of your brand.
Lastly, communication units are your communication strategy. So, it is the language that you use to speak with different categories of customers.
Getting results: start small, scale and measure success
Adaptive marketing is a powerful tool that can give you impressive results.
Big corporations like IBM, Nike, and Amazon use it. Amazon provides another example of how your buying experience shapes the content you are offered on the website. They offer product recommendations of what you might be interested in buying, based on your previous purchases. The more orders you place on Amazon, the more accurate their predictions become.
Real estate agencies use adaptive marketing as well. Zillow adapts its website content based on a customer’s financial opportunities. They show the offers from the price category a customer most likely can afford based on their IP address, and therefore the district they are from.
These companies have immense financial resources, so they can afford to use all the big-data tools for data mining and building predictive models.
Smaller companies also use adaptive marketing when they want to offer personalized messages to their clients to increase conversion rates and profits. You can set up this process manually, on a smaller scale, without any special tools.
Try drawing an experience map, segment your target audience manually, and create several personalized messages. Test and measure your results. Then, scale and try automating the process.
You can automate it with any tool you like. Hire a developer to build a program that will work for your product or buy a ready-made solution. You can even get a free, open-source tool. Any of these options will do.
Adaptive marketing isn’t simple. It requires a completely new, personalized approach to communicating with the client. But if you get the hang of it, you can increase your company’s efficiency and achieve better financial outcomes.
It can take months or even years to set up the process. At Template Monster, we have been working on it for 5 years, so that one day we could say, “We are using adaptive marketing to its fullest.” We’re not yet close to our final goal, but we have made significant progress.
Besides, we’re trying to adopt the same approach at Weblium, my new website building project. It’s a perfect launchpad for any new ideas, as we can implement new ideas on the fly and then scale them as the project evolves.
Many companies start small. Most have limited budgets and technical resources, but, like them, you should use all the opportunities that come your way to keep up with the times. Soon, companies that use adaptive marketing will outperform everyone else, because it’s effective in the most practical sense of the word.